First, a short summary on the terminology:

- in the cschem data model atomic drawing objects (such as lines) don't 
have attributes, only groups have, and it's very easy to make groups

- a net is a group with role set to "wire-net"

- any drawing object can participate in a wire-net; sch-rnd code tries to 
handle lines in smart ways and the wire tool draws wires

- there are no "connections" within wirenet groups, having the drawing 
objects in the same wirenet group simply means they are all part of the 
same network; of course it's bad practice to create wirenet groups with 
disjunct/disconnected segments, and sch-rnd convenience tools do 
everything to avoid that (but you can do that with manual group conversion 
of course)

- there are "junctions", which are not special objects or dedicated object 
type in the data model, just graphic; a junction is when two wires of the 
same wirenet cross or three wires meet in a point

- there is a "connection" object defined in the data model: it indicates 
galvanic connections between 2 or more objects of _different_ groups; 
most typical case is when you connect a wire to a terminal: you really 
connect a line of a wirenet group to a line of a terminal group, and 
sch-rnd will automatically generate a connection object for that. 

This was just a short summary of basic terms; if you want to learn more 
about how these work in the data model, you can find the detailed spect in 
chapter 2 and 3 at http://repo.hu/projects/cschem/design/


Things to test:


1. drawing wirenets

We have a dedicated tool for this, the wire tool. It draws 1 or 2 line 
segments at once. Just like in pcb-rnd the preferred refraction style can 
be changed using the / key or more temporarily by holding down shift while 
moving the cursor.

If newly drawn wire(s)' endpoint hits to an existing wirenet, the newly 
drawn wires simply extend that existing wirenet. If there's no endpoint 
overlap, a new, anonymous wirenet is created. If a newly drawn wire line 
crosses an existing wirenet line but the new line's endpoint doesn't hit 
it, so it's a "+" style crossing, it is assumed no connection is intended.

There are all sort of nasty corner cases, like when you connect two 
existing wirenets with a new wire. Please test whatever you can think of.

There are two ways to verify if the code did what you wanted:

- if you click a wire line, it is selected; if you click it again, the 
whole wirenet group is selected; this makes it easier to see what belongs 
together

- the tree view dialog (accessible from right click context menu over the 
wire line)


2. junction graphics

Please especially test if junction dots are placed and removed as 
described above, as expected.

3. connections

You can place a terminal using {p t}. It's not necessary for testing this, 
but if you want to simulate real use patterns more closely, you can create 
a symbol group from multiple terminals and ordinary drawing objects 
(lines, arcs, polygons).

If you draw a wire that intersects with a terminal, a connection is made 
between the two objects. You can see the connection object in the tree 
view or as bright green graphics if you turn on the "connections" layer.

Note: when determinign the user intention on drawing a new wire 
intersecting an exisitng wire, only endpoint match is considered intent of 
connection. However, terminals are "hot": any intersection of 
wire-terminal will cause a connection to be made.

Please test that connections are always made and removed as terminal-wire 
intersections suggest.